The coupling of RF (radio frequency) circuits often causes signal reflection. This reflected signal is captured and sent to a “dummy load,” a resistive cell. Normally, RF systems, such as those used in communications, are specifically designed to eliminate or at least minimize signal reflectivity. However, in non-communication RF systems, such as those employed in particle accelerators, reflectivity can be a very serious problem. These systems often have a reflectivity greater than about 80% and as high as about 99%, resulting in large RF energy transfer inefficiencies.
Particle accelerators such as ATLAS (Argonne Tandem-Linac Accelerator System) at Argonne National Laboratory use superconducting resonators. These superconducting resonators have a very high Q (quality factor), and therefore have a very narrow frequency bandwidth. Therefore, careful attention must be made to ensure the resonator resonance frequency is at the frequency of the RF Energy Source (RF drive frequency) supplied to the superconducting resonator. Careful attention must also be made to ensure that the resonator RF field is in phase with the RF Energy Source. For example, the resonators used in ATLAS at Argonne National Laboratory have a loaded Q on the order of 107, and a resonance frequency of 97 MHz plus or minus only a few hertz. Frequencies outside of the resonance frequency have little to no effect.
Unfortunately, the resonance frequency of the superconducting resonators is continuously altered by factors such as cryogenic pressure variations, background mechanical vibrations known as microphonics, and ponderomotive (force from ion movement) detuning.
One method of ensuring the resonator can efficiently utilize RF energy from the RF Energy Source is to overcouple the resonator, effectively increasing the bandwidth of the resonator. By increasing the resonator bandwidth, RF phase errors due to RE resonant frequency variations in the resonator can be reduced.
Unfortunately, when overcoupled, a significant amount of RF energy from the RF Energy Source is reflected back to the RF Energy Source. As the Reflected Energy can potentially damage the RE Energy Source, this Reflected Energy is normally routed through a circulator to a “dummy load” for dissipation. As resonators used in superconducting particle accelerators can have a reflectivity as high as 99%, a very significant amount of energy is lost due to overcoupling.
Another method of ensuring the resonator resonance frequency matches the frequency of the RE Energy Source is to continuously match the resonator frequency using a VCX (voltage controlled reactance) system, VCX systems are generally focused on achieving phase synchronization with the resonance frequency of the resonator.
VCX systems such as those used by ATLAS at Argonne National Laboratory are based on PIN diodes used to switch the superconducting resonator between high and a low frequency states chosen to bracket the resonate frequency. In the high-frequency state, the resonator RF phase processes forward relative to the phase of RF Energy Source, and in the low frequency state, backwards.
Phase control is achieved with a diode driver which switches the diodes between the two states. Within the switching period, the diodes can be turned on for a controlled time, which generally vary from 5% to 95% of the switching cycle. Modulation of the duty factor provides an effectively continuous control of the direction of phase precession, hence also the mean frequency received by the resonator. Unfortunately, VCX systems add additional complexity to the resonator and are problematic in high RF field applications. This additional complexity increases startup costs, maintenance costs, as well as operating costs. These additional components also absorb energy reducing the efficiency of the system, while generating heat.
Therefore, there exists a need for a simple, reliable, low cost, and energy efficient means of recuperating reflected RF energy. More particularly, there exist a need of supplying energy to a superconducting resonator, while minimizing ene loss and system complexity.